On Mon, Jul 23, 2012 at 5:28 PM, Jordi Gutiérrez Hermoso
<address@hidden> wrote:
On 23 July 2012 05:09, Sergey Dudoladov <
address@hidden> wrote:
> 1) Is it OK to submit the patch if "hg import" rejects it ? Or
> should I try to eliminate any conflicts?
The whole point of "hg export" is that this gives the patch metadata
that says *where* the patch is applied (i.e. at which hg revision).
Unless this is a private commit that is not in the Savannah
repository, it is then impossible for your patch to not apply (well,
impossible unless you found a way to exploit a hash collision in
sha1).
> 2) Should I attach the patch to the bug report or submit it
> separately through a patch tracker ?
Attached to the bug report is fine. I'll look at it now.
- Jordi G. H.
1) I've just added a potential patch to the bug report . Bug #36920 ; file #26264
2) I didn't get the point about 'hg export' . Basically, I don't understand the sentence beginning "Unless this is a private .... "
Please give me a clue where to find more information about it.
Probably I should go into more detail about my problem with rejections.
a) I have a few clones of an Octave repo on my machine and I use one of them specifically to play with function "help"
b) I added a few lines of code to help.m and commited it to one of my local repos.
c) After that just to make sure it really works I run modified help.m in local Octave installation
d) It worked, so I exported my tip revision into a file called help_mod.diff (this extension was used in contributor's guide)
e) As a last check, I applied this patch with "hg -import" to another local repo. The patch got rejected.
However, the sequence "hg pull + hg update" worked fine
At this point, I skimmed through HG manual and though the patch rejection must be resolved manually.
My initial question was "Is it appropriate to submit a patch that someone else will have to apply manually" ?